Update Docker image builds for torch 2.7+cu126 #2514
Merged
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
PyTorch 2.7 was just released. The default wheel from PyPI is built with CUDA toolkit 12.6, compared to 12.4 with earlier PyTorch releases. This resulted in broken tests for bitsandbytes, as the Docker image that builds bitsandbytes from source only builds with CUDA 12.4.
We'll update the Dockerfiles to use CUDA 12.6, matching the PyTorch 2.7 that is now used in the tests. This will result in building the libbitsandbytes_cuda126.so version that we need instead of libbitsandbytes_cuda124.so.
Secondly, I'm removing the multi-backend-refactor version of this workflow as we're starting to move that effort into main.
cc: @BenjaminBossan